Reflection 5

a. Prediction vs. Explanation

Sometimes it is enough to know what will happen, even if we do not fully understand why. For example, a model may predict which areas will have higher crime rates. This prediction can help police prepare resources. But if policymakers want to reduce crime in the long term, they need to understand why crime happens, such as poverty, unemployment, or lack of education.

b. Simple Causal Diagram and Confounders

In my research area, one relationship could be: Employee engagement → Organizational performance. The following diagram shows this relationship along with confounding factors.

Leadership Quality
Employee Engagement
Organizational Culture
Organizational Performance

Figure: Causal Relationship between Employee Engagement and Organizational Performance

However, there may be confounders such as leadership quality, organizational culture, and resources. These factors influence both engagement and performance.

To distinguish a causal claim from a predictive association, we need to control for these confounders using research design or statistical methods. If the relationship still holds after controlling for them, we can be more confident that engagement actually causes better performance, not just predicts it.